IMPACT OF PUBLIC INSURANCE EXPANSION ON MEDICAL CARE UTILIZATION USING QUANTILE REGRESSION

 

Health cost burden is the important issue related to the expansion in public health insurances; Medicaid and CHIP.  The newly eligible for public insurance population are expected to increase medical care utilization and to increase the health cost in aggregate. I explore the impact of public insurance expansions for children on medical care cost and utilization, especially of hospital care which is the most cost-intensive among medical cares as a precursory impact of Medicaid expansion of Affordable Care Act.  

I use a large sample of pediatric discharges from Kids’ Inpatient Dataset(KID) and primary care information from Medical Expenditure Panel Survey (MEPS) to assess the impact of public insurance expansions on medical care utilization and  hospital charge distribution.  I measure the generosity of public insurance by the fraction of eligible children population using Current Population Survey(CPS) and by income threshold rule for public insurance eligibility.   

I analyze data between 2000 and 2009 with children ages from 0 to 18. I investigate how medical care utilization composition in both inpatient care and outpatient care by health insurance status  changes as public insurance eligibility changes  to test whether efficiency gain exists in medical care utilization.  Also I analyze how the distribution of hospital charge varies by  public insurance eligibility using  quantile regression with fixed effects  and how  the expansion affect differently for each points of hospital charge distribution especially upper tail bearing large share of health cost. I tested whether children with higher medical need are more likely to move into public health insurance status.

I find that eligibility expansion increases both hospital care and primary care, with each 10 percentage-point rise in eligibility leading to a 3.2% increase in hospitalizations and 5.4 % increase in primary care usage.  In compositional change, the share of Medicaid covered hospitalization significantly increases compared to significant decrease in privately insured share.    

The quantile regression estimates differ markedly across points of hospital charge per day distribution especially  in upper and lower tails.  In general hospital charge per day generally decreases  at most quantiles as the  eligibility level rises.  However, at the very upper tail(90 percentile) and lower tail(10 percentile), the hospital charge per day increases significantly as the eligibility rises. The significant increase of hospital charge at the upper tail, implies possibility of selection into public insurance of higher medical need population.  This pattern in distribution is the most significant for public insurance covered hospitalizations. However, when I restricted to respiratory diseases which are the most prevalent hospitalization reason for children, hospital charge per day decreases over  all quantiles of distribution.  These results support  the efficiency effect of expanded  public insurance coverage.